OBJECTIVE. This study applied Rasch analysis to test four versions of the Clinical Assessment of Modes (CAM), an assessment based on Taylor’s Intentional Relationship Model: CAM–P, which assesses clients’ pretreatment preferences; CAM–E, clients’ treatment experience; CAM–T, therapists’ self-reported perspective; and CAM–O, an observer rating scale.

METHOD. The CAM–P was administered to 63 inpatients. The CAM–E was administered to 110 inpatients and outpatients. Trained raters rated therapists’ modes with 59 inpatients and outpatients on the CAM–O. The CAM–T was administered to 38 therapists. Analyses of reliability and validity were conducted.

RESULTS. The CAM demonstrated adequate construct validity. All versions showed acceptable internal consistency and unidimensionality within each of the subscales. Disorder between the 5 points on the ordinal rating scale was found for the client measures (CAM–P, CAM–E) and was resolved by modifying the ratings to encompass a 4-point scale.

CONCLUSION. The four CAM versions are reliable and valid measures of therapeutic communication in rehabilitation.